Azidothymidine in combination with 5-fluorouracil in human colorectal cell lines: In vitro synergistic cytotoxicity and dna-induced strand-breaks

European~ounzal ofCancerVol. 32A, No. 7, pp. 1219-1226, 1996 Copyright 0 1996 Elsewer Science Ltd. All rights reserved Printed in Great Bntain 0959-8049196 515.00+0.00

Pergamon

MI: SO959-8049(96)00018-4

Original Paper Azidothymidine in Combination with 5Fluorouracil in Human Colorectal Cell Lines: In Vz’troSynergistic Cytotoxicity and DNA-induced Strand-breaks M. Andreuccetti,’

‘Dipartimento Medica,

G. Allegrini,2 A. Antonuzzo,2 G. Malvaldi,’ P.F. Conte,2 R. Danesi,3 M. Del Tacca4 and A. Falcone2

di Biomedicina, Ospedale

Perfezionamento

Sezione

S. Chiara,

S. Anna,

di Patologia

Via Roma,

Generale,

UniversitP

67, 56 100 Piss; 3Scuola

56 100 Piss; and 41stituto di Farmacologia Italy

di Pisa, 56100

Superiore

Medica,

di Studi

Universith

Pisa; ’ U.O.

Oncologia

Universitari e di di Piss, 56100 Piss,

The in vitro cytotoxicity of the combination of azidothymidine (AZT) and 5-fluorouracil (S-FU) against the human colorectal cancer cells SW-480, SW-620 and COLO-320DM was evaluated. The cytotoxic effects of 5-FU and AZT were determined by the assay using 2,3-bis(2-methoxy-4-nitro-5-sulfophenil)2H-tetrazolium-5-carboxanilide inner salt @XT), while drug-induced DNA strand-breaks were measured using a fluorometric analysis of DNA unwinding. After an exposure of 72 h, 5-FU and AZT induced a dose-dependent cytotoxicity against each cell line. The addition of 3,10 and 30 PM AZT to various concentrations of 5-FU, as well as the addition of 0.5,l and 3 pm 5-FU to various concentrations of AZT, resulted in an enhanced cytotoxic effect. Isobologram analysis and the combination index (CI) method demonstrated that the interaction between 5-FU and AZT was clearly synergistic in each cell line, except for the 30% level of effect in SW-620, where borderline synergism was observed. The evaluation of DNA strand-breaks after an exposure of 16 h to 5-FU, AZT or 5-FU + AZT demonstrated that the 5-FU + AZT combination produced the greatest DNA damage, and that this interaction was synergistic in each cell line. In conclusion, our study supports the evidence that the potential antitumour activity of AZT can be modulated by combining it with agents which inhibit thymidylate (dTMP) formation, such as 5-FU, and that the increased cytotoxicity is related to enhanced DNA damage. These findings should encourage firther experimental and clinical studies of the potential use of AZT in combination with inhibitors of de novo dTMP synthesis. Copyright 0 1996 Elsevier Science Ltd Key words: azidothymidine, strand-breaks EurJCancer,

5-fluorouracil,

Vol. 32A, No. 7, pp. 1219-1226,

colorectal cancer cell lines, synergistic cytotoxicity, DNA 1996

INTRODUCTION is a thymidine analogue that is converted intracellularly to AZT triphosphate (AZTTP) by the sequential action of mammalian thymidine kinase (TK), thymidylate kinase and nucleoside diphosphate kinase, and is ultimately incorporated into DNA where it blocks chain elongation [l]. AZT was originally developed as an anticancer agent, but early studies did not find significant cytotoxic activity and its further development as an anticancer agent

AZIDOTHYMIDINE (AZT)

Correspondence to A. Falcone. Received 25 Jul. 1995; revised 30 Sep. 1995; accepted 4

Dec. 1995.

was therefore abandoned [2,3]. Instead, AZT has been found to show important in vitro activity against HIV and clinical utility in patients with AIDS and AIDS-related diseases [4,5]. AZT is, indeed, a more favourable substrate for HIV reverse transcriptase than for human DNA polymerase and is, therefore, preferentially incorporated into viral DNA [ 11. More recently, experimental studies by Brunetti and colleagues [6] and Tosi and colleagues [7] have shown that the incorporation of AZT into cellular DNA of the human colorectal adenocarcinoma cell line, HCT-8, can be significantly enhanced when AZT is combined with agents that inhibit the de YZOZIO synthesis of thymidylate (dTMP), such as

1219

1220

M. Andreuccetti et al.

5-fluorouracil (5-FU) or methotrexate (MTX), thus resulting in enhanced AZT-induced DNA strand-breaks and cytotoxicity. The mechanism by which 5-FU and MTX enhance AZT cytotoxicity seems to be dependent on their ability to deplete intracellular thymidine triphosphate pools (dTTP) that consequently facilitate the utilisation of AZTTP in DNA synthesis [6-81. Of interest, these studies have demonstrated that this effect is relatively specific for tumour cells, and this might reflect differences in thymidine salvage ability and transport between colorectal cancer cells and normal cells [6-81. In fact, colorectal cancer cells usually have an elevated TK activity [9-121, and, therefore, a high capacity to activate AZT. In addition to this, cancer cells lack a Na+-dependent concentrative nucleoside transport system that has been demonstrated in a variety of normal cells and which does not recognise AZT as a substrate [13-l 71. These systems, by concentrating thymidine intracellularly, may protect normal cells from the action of AZTIP for competition between AZTTP and dTTP. In this study, we extended our research on 5-FU and AZT to three human colorectal cancer cell lines to validate observations previously reported in HCT-8 cells and, in particular, to evaluate in vitro interactions between the two agents with regard to cytotoxicity and DNA damage. MATERLALS AND METHODS Drugs and chemicals AZT was kindly provided by Burroughs Wellcome Co. (Beckenham, U.K.) and 5-FU was purchased from Sigma Chemical Co. (St. Louis, Missouri, U.S.A.). The tetrazolium reagent 2,3-bis(2-methoxy-4-nitro+sulfophenil)-2H-tetrazolium-5-carboxanilide inner salt (XTT) was purchased from Polyscience (Warrington, Pennsylvania, U.S.A.) and prepared at a concentration of 1 mg/ml in prewarmed RPMI-1640 medium before use. Phenazine methosulphate (PMS) was purchased from Sigma and prepared at a concentration of 5 mM (1.53 mg/ml) in phosphate-buffered saline (PBS) and stored at 4°C for a maximum of 3 months. All other chemicals and solvents were obtained from Sigma Chemical Co. Cell line and culture conditions The human colorectal cancer cell lines SW-480, SW-620 and COLO-320DM were used in this study. These cell lines have been previously characterised [18-201 and were purchased from American Type Culture Collection (Rockville, Maryland, U.S.A.). Exponentially growing cells were grown in RPMI-1640 or Leibovitz’s L-15 medium supplemented with 10% fetal bovine serum (FBS) and 2 mM L-glutamine, and maintained in a humidified incubator at 37°C in an atmosphere of 5% CO,. In vitro evaluation of cytotoxicity The XTT assay was used to determine the effect of 5-FU and AZT on colorectal cancer cell proliferation and was performed as previously described [21]. Briefly, single cell suspensions were prepared by trypsinisation, and approximately 2000 cells, suspended in 100 ~1 of complete medium, were seeded into each well of a 96-well microtitre plate, and incubated at 37°C with 5% CO,. After 24 h, 100 ~1 of complete medium containing 5-FU and/or AZT was added to treat the cells for 72 h. Drug concentrations tested ranged between 0.5 and 100 +M for 5-FU and between 1 and 200 FM for AZT; in the combination studies, only one of the

two agents was tested at three different concentrations that had achieved a percentage growth inhibition between 5 and 30%. Fifty microlitres of medium containing 50 pg of XTT and 0.38 kg of PMS was then added to each well and plates were incubated for an additional 3-4 h. The absorbance was measured at 450 nm with a microplate reader (Lab System Instruments, Helsinki, Finland). Cell growth inhibition was expressed as the percentage absorbance of untreated cells, and the 50% inhibitory concentration of cell growth (ICY,,) was calculated by non-linear regression analysis. Fluorometni detection of DNA strand-breaks The fluorometric analysis of DNA unwinding was performed as previously reported [22]. In fact, when DNA is exposed to moderately alkaline solutions, the rate of DNA unwinding can be used as a sensitive measure of strand-breaks because the level of unwinding increases with increasing DNA strand-breaks. In this method, the fluorescent dye, ethidium bromide, that binds selectively to double-stranded DNA is used, and this procedure does not, therefore, require a physical separation of single-stranded DNA from double-stranded DNA as required in previous methods. Briefly, 5-l 0 x 1O6 cells were plated in 100 mm sterile Petri dishes and exposed to 10 p,M AZT or 1 FM 5-FU or their combination. After an exposure of 16 h to the drugs, cells were trypsinised, centrifuged and the pellet resuspended in 2 ml of solution A (0.25 M myo-inositol, 10 mM sodium phosphate, 1 mM MgC&). Aliquots of this suspension (0.2 ml) were distributed into 12 disposable glass tubes, designated T, P or B in groups of four. T tubes were used to estimate the total fluorescence (double-stranded DNA + contaminants), P tubes to estimate the unwinding rate of DNA, and B tubes (blank) to estimate the contribution of components other than double-stranded DNA to the fluorescence. For this purpose, 0.2 ml of solution B (9 M urea, 10 mM NaOH, 2.5 mM cyclohexanediaminetetraacetate, 0.1% sodium dodecylsulphate (SDS)) was added to each tube and incubated at 0°C for 10 min. After this time, during which cell lysis and chromatin disruption occurred, 0.1 ml of solution C (45% of solution B and 55% 0.2 N NaOH) and 0.1 ml of solution D (60% of solution B and 40% of 0.2 N NaOH) were gently added to the P and B tubes. After further incubation at 0°C for 30 min, during which the alkali diffused into the lysate to give a final pH of approximately 12.8, the contents of the B tubes were sonicated for l-2 s to ensure rapid denaturation of the DNA in the alkaline solution. P and B tubes were further incubated at 15°C for 60 min and denaturation was then stopped by chilling the tubes to 0°C and adding 0.4 ml of solution E (1 M glucose, 14 mM 8mercaptoethanol) that lowered the pH to approximately 11 .O. The lysates were briefly sonicated to render them homogeneous, diluted with 1.5 ml of solution F (ethidium bromide, 6.7 kg/ml, 13.3 mM NaOH) and their fluorescence was read at room temperature in a Perkin-Elmer spectrofluorometer (excitation wavelength 520 nm, emission 590 nm). The T tubes differed from the P tubes in that the neutralising solution E was added before the alkakine solutions C and D, so that the DNA was never exposed to a denaturing pH. The extent of DNA unwinding after a given time of exposure of cell extracts to alkali was calculated from the fluorescence values of the T, P and B samples. The percentage of double-stranded DNA was calculated using the formula (P-B)/(T-B) x 100 where (P-B) provides an estimate of the amount of double-

AZT+5-FU

in Colorectal

Cancer

r

90

loo oAZT Alone AAZT + S-FU 0.5 pM 90 oAZT + 5-FU 1 pM

80

80 -

v 5-FU Alone x5-FU + AZT 3 FM +5-FU + AZT 10 FM *5-FU + AZT 30 pM

70 -

s 70+ $60 8 50:: 9 40-E s2 30-

6c50 40 30 -

20 -

20 -

10 -.

10 -

0.1

1221

(b) SW-480

(a) SW-480 loo

0

Cell Lines

I I

I 10

I 100

I 1000

(d) SW-620

(c) SW-620 o AZT Alone

l*r

AAZT + 5-FU 0.5 p.M oAZT + 5-FU 1 pM o AZT + 5-FU 3 pM

r5-FW x5-FU +5-FU *5-FU

Alone + AZT 3 pM + AZT 10 PM + AZT 30 &M

v 5-FU x5-FU +5-FU *5-FU

Alone + AET 3 p.M + AZT 10 p,M + AZT 30 pM

20 IO 0.1

I

(e) COLO-320

IO

loo

loo0 (f) COLO-320

DM o AZT AAZT oAZT oAZT

100-

DM

100 Alone + 5-FU 0.5 FM 90 + 5-N 1 pM 80 + 5-FU 3 pM 70 60 50 40 30 20 -

,

lo;_L\$ I

0.1

I

IO AZT (PM)

loo

loocl

0

0.1

I

IO

loo

loo0

5-FU (PM)

Figure 1. Cytotoxic effect of treatment for 72 h with axidothymidine (AZT), 5-fluorouracil (5-FU), AZT + 5-FU (0.5, 1 and 3 JLM), and S-FU + AZT (3, 10 and 30 PM) on cell survival of the human colorectal turnout cell lines SW-480, SW-620, and COLO320DM as determin ed by percentage control absorbance in the 2,3-bis(2-methoxy-4-nitro-5-sulfophenil)-2H-tetraxolium-5carboxanilide inner salt (XTT) assay. Points are the mean of triplicate experiments each done in sextuplicate, with standard deviation shown by vertical bars.

1222

M. Andreuccetti

stranded DNA remaining after exposure to alkali for 60 min, and (T-B) an estimate of the amount of double-stranded DNA in the cell extracts. Up to the point at which solution C was added, all steps were carried out under ordinary room illumination; after this point, manipulations were carried out in the dark and a covered bath was used for incubation. All solutions were kept at O”C, except solutions B and F that were stored at room temperature. Data analysis An improved isobologram method that has been previously described [23,24] was used to determine the in vitro effects of combining AZT and 5-FU. In fact, for agents with a nonlinear dose response curve, such as 5-FU and AZT, the expected additive response cannot be represented in the isobologram as a straight line, and it is necessary to determine an area called ‘envelope of additivity’ within which all responses are conceivably additive. Thus, when a data point lies within the envelope of additivity, an additive interaction is indicated; when a data point lies to the left side of the envelope, a supraadditive or synergistic interaction is indicated, and when it lies to the right side, a subadditive or antagonistic interaction is indicated. In order to determine the envelope of additivity, mode I, IIa and IIb lines were drawn. These lines were obtained by plotting rhe doses of AZT and 5-FU which give log survival values that add up to the chosen level of effect (30% and 50%): for mode I line, the increments in dose for both AZT and 5-FU starting from zero were taken; for mode IIa line, the increments in dose for AZT starring from zero and for 5-FU starting from where the effect of AZT had ended were taken; and for mode IIb line, the increments in dose for 5-FU starting from zero and for AZT starting from where the effect of 5-FU had ended were taken. The area surrounded by these three lines represents the envelope of additivity. We also used the combination index (CI), as previously described [25], to compare the cytotoxic effects of AZT, 5FU and AZT + 5-FU. The CI was defined as the sum of rhe relative concentrations (e.g. IC+, units) of each drug which yield an isoeffect (e.g. an inhibition of 0.5) when added together: CI = (concentration of AZT)/(IC value of AZT) + (concentration of 5-FU)/(IC value of 5-FU).

Table 1. Cytotoxic e#ects of azidothymidine Cell line

Level of effect (%)

(AZT),

et al.

With this method, CI < 1 indicates synergy, CI > 1 antagonism and CI = 1 additivity. Interaction between 5-FU and AZT on DNA-induced strand-breaks was analysed using a multiplicative model as previously described [26,27]. Briefly, the multiplication of the fraction of double-stranded DNA (DS-DNA) remaining after exposure to either drug alone (DS-DNA,.N or DS-DNA,zT) is compared with the observed result. If DS-DNA,_FU+AzT < (DS-DNA,_FU x DS-DNAA,,), the interaction is synergistic; if DS-DNA5_FU + AZT= (DS-DNA,., x DS-DNA,=), the interaction is additive; and if DS-DNA,_, +AzT > (DSDNA,_, x DS-DNA,,), it is antagonistic. RESULTS 5-FU and AZT showed dose-dependent cytotoxicity against the cell lines SW-480, SW-620 and COLO-320DM; xcsO concentrations were, respectively, 10.25, 7.65 and 7.17 FM for 5-FU and 155.42,94.76 and 149.80 FM for AZT after an exposure of 72 h (Figure 1, Table 1). The addition of AZT at 3, 10 and 30 PM co various concentrations of 5-FU, as well as the addition of 5-FU at 0.5, 1 and 3 PM to various concentrations of AZT, resulted in an enhanced cytotoxic effect (Figure 1, Table 1). Isobologram analysis of the q0 and lcgo for the drug combinations (Figure 2) indicated that the interaction between the two agents was clearly synergistic in most cases, except for the 30% level of effect in SW-620, where only a borderline synergism was observed (Figure 2~). This observation was confirmed by comparing the cytotoxic effects of 5-FU, AZT and 5-FU + AZT with the CI method at different levels of effect (30% and 50%); in fact, the interaction was synergistic for each cell line, though synergism was more pronounced for SW-480 and COLO-320DM (Table 2). As previous studies have suggested that 5-FU enhances AZT cytotoxicity by increasing the degree of AZTTP incorporation into DNA, thereby increasing AZT-induced DNA strand-breaks, the damage induced by 5-FU, AZT and 5FU + AZT on DNA was measured. After an exposure of 16 h to 1 FM 5-FU, only modest amounts of DNA strand-breaks were detected in the cell lines (Figure 3). AZT, 10 FM, produced a significant increase in the extent of DNA strandbreakage, but the greatest effect was obtained by combining 5-FU and AZT (Figure 3). In fact, after an exposure of 16 h to the two agents, the mean values of double-stranded DNA

5-fluorouracil(5-FU)

and AZT + S-FU against human colorectal cell lines 5-FU concentration (PM)

AZT concentration (PM) AZT alone

AZT AZT + 0.5 FM 5-FU + 1 FM 5-FU

AZT + 3 pM 5-FU

5-FU alone

5-FU +~FMAZT

5-FU +lOpMAZT

5-FU +30pMAZT

SW-480

30 50 70

57.48 155.42 ND

0.92 33.00 ND

0.59 0.99 44.77

0.53 0.88 35.22

3.39 10.25 93.88

0.29 0.48 33.49

0.28 0.46 3.77

0.26 0.43 1.52

SW-620

30 50 70

35.30 94.76 ND

11.90 31.59 110.27

1.85 9.08 28.53

0.57 0.95 18.11

3.26 7.65 45.09

1.40 3.56 18.57

0.87 2.27 8.59

0.37 1 .oo 2.34

COLO-320DM

30 50 70

37.10 149.80 ND

0.51 0.85 16.53

0.36 0.60 0.84

0.32 0.54 0.75

0.32 0.56 0.96

0.16 0.26 0.36

0.15 0.25 0.36

ND, not determined.

3.24 7.17 19.33

AZT+5FU

1223

in Colorectal Cancer Cell Lines

50% Effect level

30% Effect level (a) SW-480

(b) SW-480

(c) SW-620 1.0

(d) SW-620

0.8

0.8

0.6

0.6 2 +I

iz A

0.4

0.4

0.2

(f) COLO-320 DM

(e) COLO-320 DM 1.0

\,

‘\

0.8

0.6 iz Gl 0.4

0.2

0

0.2

0.4 AZT

0.6

0.8

1.0 AZT

Figure 2. Isobologram analysis of the cytotoxlc effects of the combination of azidothymidine (AZT) and S-fluorouracil (5FU) at 30% and 50% levels of effect ln the human colorectal carcinoma cell lines SW-480, SW-620 and COLO-320DM. (e), Mean ICY,, or lcsOvalues of three experiments each done in sextuplicate; (-), mode I line; (----), mode IIa line; (.*-**>, mode III, line.

M. Andreuccetti

1224

Table 2. Mean

et al.

combination

index (Cl)

Cell line

Treatment

30% level of effect CI Interaction

SW-480

AZT + 0.5 pM 5-FU AZT + 1 pM 5-FU AZT + 3 FM 5-FU

0.16 0.30 0.89

S S S

0.26 0.10 0.30

S S S

5-FU + 3 )LM AZT 5-FU + 10 pM AZT 5-FU + 30 pM AZT

0.14 0.34 0.60

S S S

0.06 0.11 0.43

S S S

AZT + 0.5 FM 5-FU AZT + 1 pM 5-FU AZT + 3 yM 5-FU

0.49 0.36 0.94

S S S

0.40 0.23 0.40

S S S

5-FU + 3 pM AZT 5-FU + 10 FM AZT 5-FU + 30 yM AZT

0.51 0.55 0.96

S S S

0.50 0.40 0.45

S S S

AZT + 0.5 FM 5-FU AZT + 1 pM 5-FU AZT + 3 pM 5-FU

0.17 0.32 0.93

S S S

0.07 0.14 0.42

S S S

5-FU + 3 pM AZT 5-FU + 10 pM AZT 5-FU + 30 pM AZT

0.18 0.32 0.85

S S S

0.09 0.10 0.23

S S S

SW-620

COLO-320DM

50% level of effect CI Interaction

AZT, azidothyrnidine; 5-FU, 5-fluorouracil; S, synergistic.

(a) SW-480

(b) SW-620

(c) COLO-320 DM

100

100

90

90

80

80

70

70

60

60

50

50

40

40

30

30

20

20

10

10

0

0

Figure 3. Effects of exposure for 16 h to 1 PM S-fluorouracil (SFU), 10 PM azidothymidme (AZT) or 1 PM 5-FU + 10 WM AZT on the amount of residual double-stranded DNA in SW-480, SW-620 and COLO-320DM colorectal carcinoma cell lines. Results are the mean of three experiments each done in duplicate, with standard deviation shown by vertical bars.

were 22, 21.8 and 27.4% in SW-480, SW-620 and COLO320DM, respectively, in comparison to a mean control value of approximately 95%. The effect of combining 5-FU and AZT on DNA was analysed using the multiplicative model previously described, and it was found to be synergistic for each cell line. DISCUSSION Despite recent advances in medical oncology, the treatment of metastatic colorectal cancer remains unsatisfactory. Although 5FU is considered to be the most effective drug, it induces remissions only in a minority of patients, with no

significant impact on survival [28]. Numerous experimental studies have attempted to modulate 5FU activity biochemically, but only modulation with leucovorin or MTX has produced some clinical improvement, leading to higher response rates and, in some cases, a modest survival advantage [29-311. Recent studies have suggested that a new approach to enhance the cytotoxicity of 5-FU or other dTMP synthesis inhibitors, i.e. MTX, might be to combine them with a toxic thymidine analogue, such as AZT [6,7]. In fact, under conditions of de novo dTMP synthesis inhibition, incorporation of AZT into DNA is facilitated. Interestingly, in vivo studies have demonstrated that this effect is relatively specific

AZT+5-FU

in Colorectal

for tumour cells, and reflects differences in pyrimidine salvage capacity between tumour and normal cells [6,7]. Our study confirms, in SW-480, SW-620 and COLO320DM human colorectal cancer cell lines, the previous in vitro observations for the cancer cell line HCT-8. Of interest, the interaction between 5-FU and AZT was clearly synergistic in at least two of the cell lines (SW-480 and COLO-320DM). In addition to this, our study suggests that the enhanced cytotoxicity induced by the drug combination is possibly related to an increase in DNA-induced strand-breaks that can be produced by enhanced AZT incorporation into DNA, as previously demonstrated. Recently, other investigators have reported similar results by combining AZT with de nova dTMP synthesis inhibitors. Tosi and associates [32] showed that hydroxyurea, which inhibits ribonucleotide reductase activity and, therefore, reduces deoxyribonucleotide pools, increases the incorporation of AZT into DNA and synergistically enhances its cytotoxicity in two human chronic myeloid leukaemia cell lines [32]. Moreover, Presacco and Erlichman [33] reported that ICI D1694, a new folate-based thymidylate synthase inhibitor, also enhances AZT cytotoxicity and DNA-induced strand-breaks in the bladder MGH-Ul and colorectal HCT8 tumour cell lines, but not in normal CFU-GM cells. Finally, a preliminary report by Szekeres and colleagues [34] has confirmed a synergistic cytotoxic effect of 5-FU and AZT in the human colon cancer cell lines HT-29, CCL-227, CCL228 and Caco-2. Clinical studies combining AZT with 5-FU or MTX have been initiated. Phase I studies have shown that the combination is clinically feasible and that AZT doses of up to 810 g/m’ administered over 2-24 h do not significantly enhance 5-FU- or MTX-induced toxicities. Interestingly, the drug combination increases DNA strand-breaks dose-dependently in peripheral blood-nucleated cells [35,36]. Furthermore, preliminary phase II studies demonstrate promising results for 5FU + leucovorin + AZT in metastatic colorectal cancer and for MTX + AZT in HIV-related high-grade non-Hodgkin’s lymphomas [37,38]. In conclusion, our study supports the concept that the potential antitumour activity of AZT can be modulated by combining it with agents that inhibit dTMP formation and reduce intracellular dTI;P pools, as in the case of 5-FU, and that the increased cytotoxicity is related to enhanced AZTinduced DNA damage. Because in vitro and in vivo studies indicate that this effect is relatively specific for tumour cells, the combination of AZT with agents that inhibit dTMP formation may be clinically useful and should be further evaluated.

Cancer

5. Fischl MA, Richmann DD, Grieco MH, et al. The efficacy of azidothymidine in the treatment of patients with AIDS and AIDSrelated complex. NEnglJ Med 1987,317, 185-191. 6. Brunetti I, Falcone A, Calabresi P, et al. 5-Fluorouracil enhances azidothymidine cytotoxicity: in vitro, in z&o, and biochemical studies. Cancer Res 1990, 50,4026-403 1. 7. Tosi P, Calabresi P, Goulette FA, et al. Azidothymidine-induced cytotoxicity and incorporation into DNA in the human colon tumor cell line HCT-8 is enhanced by methotrexate in vitro and in vivo. Cancer Res 1992, 52, 4069-4073. 8. Darnowski JW, Goulette FA. 3’-Azido-3’-deoxythymidine cytotoxicity and metabolism in the human colon tumor cell line HCT8. Biochem Pharmacoll994,48, 1797-1805. 9. Demon JE, Lui MS, Aoki T, et al. Enzymology of pyrimidine and carbohydrate metabolism in human colon carcinomas. Cancer Res 1982,42,1176-l 183. 10. Weber G, Lui MS, Takeda E, et al. Enzymology of human colon tumors. L.;fe Sci 1980,27,793-799. 11. Sakamoto S, Sagara T, Iwama T, et al. Increased activity of thymidine kinase isozymes in human colon polyp and carcinoma. Carcinogenesis 1985,6, 9 17-9 19. 12. Herzfeld A. Leers: MA, Greeneard 0. Human colon tumors. Enzymic and his%ogical characteristics. Cancer 1978, 42, 12801283. 13 Wohlhueter RM, Marz R, Plagemann PGW. Thymidine fransport in cultured mammalian cells. Kineric analysis, temperature dependence and specificity of the transport system. Biochim Biophys Acta 1979, 553,262-283. 14. Paterson ARP, Kolassa N, Cass CE. Transport of nucleoside drugs in animal cells. Pharmacol Ther 1981, 12, 516-536. 15. Dagnino L, Bennett LL, Paterson ARP. Concentrative transport of nucleosides in Ll210 mouse leukemia cells. Proc Am Assoc Cancer Res 1987,28, 15. 16. Belt JA, Vijayalakshmi D. Sodium-dependent nucleoside transport in mouse intestinal epithelial cells: two transporters with differing substrate specificities. Proc Am Assoc Cancer Res 1988, 29, 13. 17. Domin BA, Mahony WB, Zimmerman TP. 2’-3’-Dideoxythymidine permeation of the human erythrocyte membrane by nonfacilitated diffusion. Biochem Biophys Res Commun 1988, 154, 825-831.

18. Fogh J, Trempe Human

Tumor

USA 1985,82,7096-7100.

G. New human

tumor cell lines. In Fogh J, ed. York, Plenum Publishing,

CeZls in Vitro. New

1975,115-159. 19. Leibovitz A, Stinson JC, McCombs WB, McCoy CE, Mazur KC, Mabry ND. Classification of human colorectal adenocarcinoma cell lines. Carrcer Res 1976, 36,4562-4569. 20. Quinn LA, Moore GE, Morgan RT, Woods LK. Cell lines from human colon carcinoma with unusual cell products, double minutes, and homogeneously staining regions. Cancer Res 1979, 39,4914-4924. 21. Scudiero DA, Shoemaker RH, Paul1 KD, et al. Evaluation of a soluble retrazolimniformazan assay for cell growth and drug sensitivity in culture using human and other tumor cell lines. Cancer Res 1988,48,4827-4833. 22. Birnboim HC, Jevcak JJ. Fluorometric method for rapid detection of DNA strand breaks in human blood cells produced by low dose ofradiation. CancerRes 1981,41, 1889-1892. 23. Steel GG, Peckham MJ. Exploitable mechanisms in combined radiotherapy-chemotheapy: the concept of additivity. IntJRadiat OncolBiolPhys

1. Furman PA, Fyfe JA, St. Clair MH, et al. Phosphorylation of 3’-azido-3’-deoxythymidine and selective interaction of the Striphosphate with humax immunodeficiency virus reverse transcriptase. Proc Nat1 Acad Sci USA 1986, 83, 8333-8337. 2. Horowitz JP, Chua J, Noel MJ. Nucleosides. V. The monomesylates of l-(2’-deoxy-beta-D-1yxofuranosyl)thymidine. 3 Org Chem 1964,29,2076-2078. 3. Wilde MI, Langtry HD. Zidovudine. An update of irs pharmacodynamic and pharmacokinetic properties and therapeutic efficacy. Drugs 1993,46, 5 15-578. 4. Mitsuya H, Weinhold KJ, Furman PA, et al. 3’-Azido-3’-deoxythymidine: an antiviral agent that inhibits the infectivity and cytopathic effect of human T-lymphotropic virus type III/lymphadenopathy-associated virus in vitro. Proc Natl Acad Sci

1225

Cell Lines

1979,

5, 85-91.

24. Kano Y, Ohnuma T, Okano T, et al. Effects of vincristine in combination with methotrexate and other anti-tumor agents in human lymphoblastic leukemia cells in culture. Cancer Res 1988, 48,351-356. 25. Tsai CM, Hsiao SH, Frey CM, et al. Combination cytoroxicity effects of cis-diamminodichloroplatinum and 5-fluorouracil with and without leucovorin against human non-small cell cancer cell lines. Cancer Res 1993, 53, 1079-1084. 26. Falcone A, Danesi R, Zaccaro L, et al. Synergistic antiproliferative activity of suramin and a2A-interferon against human colorectal adenocarcinoma cell lines: in vitro studies. Eur J Cancer 1994, 30A, 516-520. 27. Sayers TJ, Wiltrout TA, McCormick K, Husted C, Wiltrout RH. Antitumor effects of a-interferon and y-interferon on a murine renal cancer (Renca) in vitro and in uivo. Cancer Res 1990, 50, 5414-5420.

1226

M. Andreuccetti

28. Ahlgren JD. Colorectal cancer: chemotherapy. In Ahlgren JD, Macdonald JS, eds. Gastrointestinal Oncology. Philadelphia, J.B. Lippincott, 1992, 339-357. 29. Advanced Colorectal Cancer Meta-Analysis Project. Meta-analysis of randomized trials testing the biochemical modulation of fluorouracil by methotrexate in metastatic colorectal cancer. 3 Clin Oncoll994,12, 960-969. 30. Advanced Colorectal Cancer Meta-Analysis Project. Modulation of fluorouracil by leucovorin in patients with advanced colorectal cancer: evidence in terms of response rate. 3 Clin Oncol1992, 10, 896-903.

3 1. Sotos GA, Grogan L, Allegra J. Preclinical and clinical aspects of biomodulation of 5fluorouracil. Cancer Treat Rev 1994, 20, 1 l-49. 32. Tosi P, Visani G, Ottaviani E, er al. Hydroxyurea enhances 3’azido-3’-deoxythymidine (AZT) cytotoxicity in human chronic myeloid leukemia models. Eur3 HaematoZl994, 52, 291-295. 33. Presacco J, Erlichman C. Combination studies with 3’-azido3’-deoxythymidine plus ICI D1694: cytotoxic and biochemical effects. Eiochem Pharmacoll993,46, 1989-1992.

et al.

34. Szekeres T, Findenig G, Fritzer M, Mader R, Schon HJ. Synergistic cytotoxic effects of 5-fluorouracil with AZT in human colon carcinoma cells. Proc Am Assoc Cancer Res 1995, 36,297. 35. Posner MR, Damowski JW, Weitberg AB, et al. High-dose intravenous zidovudine with 5-fluorouracil and leucovorin. A phase I trial. Cancer 1992,70,2929-2934. 36. Browne MJ, Beitz J, Clark FJ, et al. A phase I study of zidovudine combined with methotrexate in patients with advanced cancer. Proc Am Sac Clin Oncol1993, 12, 163. 37. Falcone A, Dargenio F, Brunetti I, er al. Intravenous azidothymidine in combination with 5-fluorouracil and I-folinic acid: a phase I-II study in metastatic colorectal carcinoma. Proc Am Sot Clin Oncoll994, 13, 198. 38. Tosi P, Gherlinzoni F, Visani G, et al. Zidovudine as an antineoplastic agent in the treatment of HIV related high-grade nonHodgkin lymphoma. Blood 1994,84 (Suppl. l), 518a. Acknowledgement-The

from the Associazione M.U.R.S.T. 60%.

work was partially supported by a grant Italiana per la Ricerca sul Cancro and